Plane-Wave Decomposition with Aliasing Cancellation for Binaural Sound Reproduction

D. L. Alon and B. Rafaely (Ben-Gurion University of the Negev)
International Audio Engineering Society (AES) Convention, New York, USA, October 2015
[showhide type=”Abstract”] Abstract: Spherical microphone arrays are used for capturing three-dimensional sound fields, from which binaural signals can be obtained. Plane-wave decomposition of the sound field is typically employed in the first stage of the processing. However, with practical arrays the upper operating frequency is limited by spatial aliasing. In this paper a measure of plane-wave decomposition error is formulated to highlight the problem of spatial aliasing. A novel method for plane-wave decomposition at frequencies that are typically considered above the maximal operating frequency is then presented, based on the minimization of aliasing error. The mathematical analysis is complemented by a simulation study and by a preliminary listening experiment. Results show a clear perceptual improvement when aliasing-cancellation is applied to aliased binaural signals, indicating that the proposed method can be used to extend the bandwidth of binaural signals rendered from microphone array recordings. [/showhide]
Link to paper

Binaural Reproduction of Finite Difference Simulations using Spherical Array Processing

J. Sheaffer (Ben-Gurion University of the Negev), M. van Walstijn (Queens University, Belfast), B. Rafaely (Ben-Gurion University of the Negev) and K. Kowalczyk (AGH University of Science and Technology, Krakow)
IEEE Transactions on Audio, Speech and Language Processing, vol. 23, no. 12, pp. 2125-2135, December 2015
[showhide type=”Abstract”] Abstract: Due to its efficiency and simplicity, the finite-difference time-domain method is becoming a popular choice for solving wideband, transient problems in various fields of acoustics. So far, the issue of extracting a binaural response from finite difference simulations has only been discussed in the context of embedding a listener geometry in the grid. In this paper, we propose and study a method for binaural response rendering based on a spatial decomposition of the sound field. The finite difference grid is locally sampled using a volumetric array of receivers, from which a plane wave density function is computed and integrated with free-field head related transfer functions, in the spherical harmonics domain. The volumetric array is studied in terms of numerical robustness and spatial aliasing. Analytic formulas that predict the performance of the array are developed, facilitating spatial resolution analysis and numerical binaural response analysis for a number of finite difference schemes. Particular emphasis is placed on the effects of numerical dispersion on array processing and on the resulting binaural responses. Our method is compared to a binaural simulation based on the image method. Results indicate good spatial and temporal agreement between the two methods. [/showhide]

On Natural Robot Movements for Enriching Acoustic Information

S. Bodiroza (Humboldt Universität zu Berlin), V. Tourbabin (Ben-Gurion University of the Negev), G. Schillaci, J. Sheaffer (Ben-Gurion University of the Negev), V. Hafner (Humboldt Universität zu Berlin) and B. Rafaely (Ben-Gurion University of the Negev)
Israeli Conference on Robotics (ICR 2016), Herzelia, Israel, April 2016
[showhide type=”Abstract”] Abstract: The way robots move affects both the naturalness of human-robot interaction and the quality of their own sensory perception. The effect of robot head motion on these aspects has been jointly analyzed in the current work. The analysis was based on a human-robot interaction experiment and an acoustic simulation, both of which were performed using the same robot motion trajectories. The human-robot interaction experiment measured how the change in underlying motion parameters, namely velocity, velocity profile, and total angle of head rotation, affects the way humans perceive these movements. In the acoustic analysis, the effect of the same motion parameters on the ability to enhance auditory information acquired by the robot was studied. The results show that auditory aspects and human-robot interaction aspects are largely independent. In particular, human perception of the robot was found to be dependent on the velocity profile of the robot’s head, while the robot’s auditory performance was found to be dependent mainly on the velocity and the total head rotation angle. These findings can foster development of guidelines for the design of humanoid robot motion while taking both aspects, the naturalness of interaction and the acoustic performance, into account. [/showhide]
Copyright Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author’s copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.
Paper: Paper_ICR_2016_UBER_SB

Optimal Design of Microphone Array for Humanoid-Robot Audition

V. Tourbabin and B. Rafaely (Ben-Gurion University of the Negev)
Israeli Conference on Robotics (ICR 2016), Herzelia, Israel, April 2016
[showhide type=”Abstract”] Abstract: One of the important components of a humanoid robot is its auditory system. The system is mainly aimed at increasing the robot’s awareness of its surroundings and at enabling natural human-robot interaction using speech. The auditory system is usually based on a microphone array which constitutes the front end for sound acquisition. Configuration of this array plays a central role in the performance of the system as a whole. There are robot-audition related publications concerned with the optimization of the array configuration for enhancing the spatial information acquired by the array [1] and for improving the sound localisation performance [2]. However, spatial aliasing, which is one of the major problems in array design, remains largely untreated in the humanoid robot audition literature. The current work presents a method for microphone positioning optimisation that extends the aliasing-free frequency range of the array. The method can be used to complement the existing techniques for aliasing cancellation by signal processing [3]. The efficacy of the proposed method to reduce aliasing is demonstrated by showing a significant performance improvement, when compared to using the efficient nearly-uniform microphone distribution. The proposed method is applied to the design of a new 12-microphone array for the NAO robot. This design is subject to the real constraints on microphone positioning due to the robot’s cameras, loudspeakers, and other components. An initial evaluation of the new prototype head manufactured by Aldebaran Robotics according to the design is discussed and an example is presented showing its efficacy for reducing the speech-recognition error rates in real-world environments. [/showhide]
Copyright Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author’s copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.
Paper: Paper_ICR_2016_BGU_VT

Design of Pseudo-Spherical Microphone Array with Extended Frequency Range for Robot Audition

V. Tourbabin and B. Rafaely (Ben-Gurion University of the Negev)
Jahrestagung für Akustik DAGA 2016, Aachen, Germany, March 14-17, 2016
[showhide type=”Abstract”] Abstract: Microphone arrays constitute the front end for sound acquisition in the auditory system of most humanoid robots. Their design and performance therefore play a central role in robot audition. Although some previous studies are concerned with the optimization of microphone placement for robot audition, spatial aliasing, constituting a major challenge in array design, has not been studied extensively for this application. Spatial aliasing is approached in this paper using a spherical harmonics (SH) framework. A method for microphone positioning that extends the aliasing-free frequency range of low SH orders is developed, and validated by simulation for the design of a microphone array on the head of the humanoid NAO. Aliasing level is significantly reduced compared to the efficient nearly-uniform microphone distribution. The proposed method was employed in the implementation of a 12-microphone array for NAO. [/showhide]
Copyright Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author’s copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.
Paper: Paper_DAGA_2016_BGU_VT

 

Beamforming with Optimal Aliasing Cancellation in Spherical Microphone Arrays

D. L. Alon and B. Rafaely (Ben-Gurion University of the Negev)
IEEE Transactions on Audio, Speech and Language Processing, vol. 24, no. 1,  p. 196-210. January 2016.
[showhide type=”Abstract”] Abstract: Spherical microphone arrays facilitate three-dimensional processing and analysis of sound fields in applications such as music recording, beamforming and room acoustics. The frequency bandwidth of operation is constrained by the array configuration. At high frequencies, spatial aliasing leads to sidelobes in the array beam pattern, which limits array performance. Previous studies proposed increasing the number of microphones or changing other characteristics of the array configuration to reduce the effect of aliasing. In this paper we present a method to design beamformers that overcome the effect of spatial aliasing by suppressing the undesired side-lobes through signal processing without physically modifying the configuration of the array.
This is achieved by modeling the expected aliasing pattern in a maximum-directivity beamformer design, leading to a higher directivity index at frequencies previously considered to be out of the operating bandwidth, thereby extending the microphone array frequency range of operation. Aliasing cancellation is then extended to other beamformers. A simulation example with a 32-element spherical microphone array illustrates the performance of the proposed method. An experimental example validates the theoretical results in practice. [/showhide]
Copyright Notice ©2016 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
Paper: Paper_TASLP_2016_BGU_DA

Exploration Behaviours, Body Representations and Simulation Processes for the Development of Cognition in Artificial Agents

G. Schillaci, V. V. Hafner (Humboldt Universität zu Berlin) and B. Lara (Universidad Autónoma del Estado de Morelos)

Frontiers in Robotics and AI, section Humanoid Robotics, Dec, 2016
[showhide type=”Abstract”] Abstract: Sensorimotor control and learning are fundamental prerequisites for cognitive development in humans and animals. Evidence from behavioural sciences and neuroscience suggests that motor and brain development are strongly intertwined with the experiential process of exploration, where internal body representations are formed and maintained over time. In order to guide our movements, our brain must hold an internal model of our body and constantly monitor its configuration state. How can sensorimotor control enable the development of more complex cognitive and motor capabilities? Although a clear answer has still not been found for this question, several studies suggest that processes of mental simulation of action-perception loops are likely to be executed in our brain and are dependent on internal body representations. Therefore, the capability to re-enact sensorimotor experience might represent a key mechanism behind the implementation of higher cognitive capabilities, such as behaviour recognition, arbitration and imitation, sense of agency and self-other distinction. This work is mainly addressed to researchers in autonomous motor and mental development for artificial agents. In particular, it aims at gathering the latest developments in the studies on exploration behaviours, internal body representations, and processes of sensorimotor simulations. Relevant studies in human and animal sciences are discussed and a parallel to similar investigations in robotics is presented. [/showhide]
Copyright Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author’s copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.
Paper: Paper_FRONTIERS_2016_UBER_GS

On the Sense of Agency and of Object Permanence in Robots

S. Bechtle (Bernstein Center for Computational
Neuroscience), G. Schillaci, and V. V. Hafner (Humboldt Universität zu Berlin)

Joint IEEE International Conference on Development and Learning and on Epigenetic Robotics (ICDL-EPIROB), Paris, France, Sept. 19-22, 2016
[showhide type=”Abstract”] Abstract: This work investigates the development of the sense of object permanence in humanoid robots. Based on findings from developmental psychology and from neuroscience, we link the mechanisms behind the development of the sense of object permanence to those behind the development of sense of agency and to processes of internal simulation of sensory activity. In this paper, we present two experiments. First, a humanoid robot has to learn the forward relationship between its movements and their sensory consequences perceived from the visual input. In particular, we implement a self-monitoring mechanism that allows the robot to distinguish between self-generated movements and those generated by external events. In a second experiment, once having learned this mapping, we exploit the self-monitoring mechanism to suppress the predicted visual consequences of intended movements. We speculate that this process can allow for the development of the sense of object permanence. We will show that, using these predictions, the robot maintains an enhanced simulated image where an object occluded by the movement of the robot arm is still visible, due to sensory attenuation processes.[/showhide]
Copyright Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author’s copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.
Paper: Paper_EPIROB_2016_UBER_SB

Using Proprioceptive Information for the Development of Robot Body Representations

I. Guertel (Bernstein Center for Computational Neuroscience), G. Schillaci, and V. V. Hafner (Humboldt Universität zu Berlin)

Joint IEEE International Conference on Development and Learning and on Epigenetic Robotics (ICDL-EPIROB), Paris, France, Sept. 19-22, 2016
[showhide type=”Abstract”] Abstract: As part of the attempt to improve a robot’s flexibility and adaptation by adopting biologically inspired developmental methods, we trained a multilayer perceptron model (MLP) to develop body representations of a humanoid robot using proprioceptive and motor information. The information used were the left arm joint positions, the motor commands and the electric currents applied to these joints. By babbling its left arm, that is by executing a self-exploration behaviour, the robot gathered sensorimotor information for training the model. Once having learned the relation between these different modalities, the model can be used for running predictive processes. We present our first training results and discuss further research possibilities. [/showhide]
Copyright Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author’s copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.
Paper: Paper_EPIROB_2016_UBER_IG

How Do I Sound Like? Forward Models for Robot Ego-Noise Prediction

A. Pico, G. Schillaci, V. V. Hafner (Humboldt Universität zu Berlin) and B. Lara (Universidad Autónoma del Estado de Morelos)

Frontiers in Robotics and AI, section Humanoid Robotics, June 30, 2016,
[showhide type=”Abstract”] Abstract: How do robots sound like? Robot ego-noise, that is the sound produced by a robot while moving around, is an important factor that can affect the way an artificial agent perceives the environment and interacts with it. In robot audition, for example, ego-noise is usually addressed due to its effects on the quality of the auditory input signal, as it can severely impact the performance of processes such as speech recognition. Nonetheless, robot ego-noise can carry out very useful information about the robot embodiment or about the external environment. In this study, we present a mechanism for
learning and for predicting the auditory consequences of self-generated
movements on a custom robotic platform. We show two experiments based on a computational model capable of performing forward predictions. First, we demonstrate that the system can classify motor behaviours by comparing the noise they produce with that of simulated actions. Thus, we show that, by using similar processes, the robot can detect unexpected environmental conditions, such as changes in the inclination of the surface it is walking on. [/showhide]
Copyright Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author’s copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.
Paper: Paper_EPIROB_2016_UBER_AP